RESEARCH ARTICLE Open Access

Pre-operative factors correlated witharthroscopic reparability of large-to-massiverotator cuff tearsVanasiri Kuptniratsaikul, Thongchai Laohathaimongkol, Vantawat Umprai, Chuenrutai Yeekianand Niti Prasathaporn*

Abstract

Background: The purpose of this study is to determine the pre-operative factors that are associated withreparability of the large-sized and massive rotator cuff tears.

Methods: Sixty-six patients were included in this prognostic study. Demographic data, radiographic and MRIparameters were collected. Arthroscopic rotator cuff repair was performed for all included patient. Complete rotatorcuff repair was achieved when the tendon covered up at least 50% of the anatomical footprint. The receiver operatingcharacteristic (ROC) curve was analysed to define the cut-off level of each significant factor.

Results: Eleven large-sized rotator cuff tears and fifty-five massive rotator cuff tears were defined from MRI. Fifty-fourpatients were in the complete repair group, and twelve patients were in the partial repair group. The mean durationbetween MRI and surgery of 5.5 weeks. Reparability was correlated with age, mediolateral (ML) and anteroposterior (AP)tear size, rotator cuff arthropathy, superior migration of humeral head, fatty infiltration and atrophy of the supraspinatusmuscle, and fatty infiltration of infraspinatus muscle (p < 0.05). The ROC curve defined a cut-off level of each predictingfactor which included age of ≥65 years, mediolateral tear size of ≥36 mm, anteroposterior tear size of ≥22mm, Hamada’s rotator cuff arthropathy of ≥class2, acromiohumeral interval of ≥6 mm, ≥stage3 supraspinatusfatty infiltration, the presence of supraspinatus muscle atrophy, and ≥ stage1 infraspinatus fatty infiltration. Inmultivariated regression analysis, age, acromiohumeral interval, and anteroposterior tear size were statisticallyassociated with the reparability. The intra- and inter-observer reliabilities were moderate to excellent.

Conclusion: Age, ML tear size, AP tear size, rotator cuff arthropathy, superior migration of humeral head, fattyinfiltration of supraspinatus and infraspinatus muscles and supraspinatus muscle atrophy all correlate with reparabilityof large to massive rotator cuff tear.

Keywords: Rotator cuff tear, Reparability, Prognostic factors

BackgroundRotator cuff tear is a common cause of shoulder pain inthe general population [1, 2]. There are various surgicalapproaches available ranging from rotator cuff repair toother salvage procedures such as reverse total shoulderarthroplasty. Arthroscopic rotator cuff repair is the mostcommon procedure for treating large and massive rota-tor cuff tear as it generally gives satisfactory results inmost patients. However, 30% of the cuff tears fail to

cover 50% of their own footprint after the surgery [3].We consider these tears as partial repaired rotator cufftears which may lead to chronic shoulder pain or func-tional deficit. There are many treatment options avail-able for symptomatic partial repaired rotator cuff tearssuch as arthroscopic debridement, arthroscopic bicepsaugmentation, arthroscopic subacromial spacer, pectora-lis or latissimus dorsi tendon transfer, and reversed totalshoulder replacement.Other studies have found that post-operative clinical

outcomes are associated with age, operative time, andstructural factors, including tear size, anteroposterior

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

* Correspondence: nprasathaporn@gmail.comDepartment of Orthopaedics, Queen Savang Vadhana Memorial Hospital,Chonburi, Thailand

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 https://doi.org/10.1186/s12891-019-2485-4

and mediolateral tear lengths, tear thickness, proximal mi-gration of humeral head, muscle atrophy, fatty infiltration,and reparability [4–7]. Although a partial repair improvesthe functional result postoperatively, a complete repair, onthe other hand, shows a statistically significant superiorfunctional improvement [8]. Other studies found that tearsize, rotator cuff muscle atrophy and rotator cuff musclefatty infiltration might be associated with reparability oflarge to massive rotator cuff tear [9–12]. These factors arecrucial for creating the guidelines of managing large-sizedand massive rotator cuff tear. The purpose of our study isto explore pre-operative clinical, radiographic and mag-netic resonance imaging (MRI) parameters, which corre-lated with arthroscopic reparability of large-to-massiverotator cuff tear.

MethodsWe conducted a retrospective study on prospectivelycollected data from January 2013 to April 2017. All pa-tients with large-sized or massive rotator cuff tears,which were determined by pre-operative MRI, were in-cluded in this study. Our institute classified the rotatorcuff tears with modified Millstein and Snyder classifica-tion [13, 14]. A large-sized rotator cuff tear was definedas a complete supraspinatus tendon tear. Massive rotatorcuff tear was defined as at least two tendons involve-ment. The tear size was confirmed under arthroscopicexamination. Patients with rotator cuff re-rupture orwith contraindications for surgery were excluded. Stand-ard plain radiographs and MRI were performed in allpatients.Baseline characteristics including age, gender, duration

of symptoms, and traumatic event were collected. Fur-thermore, radiographic parameters from plain X-ray andMRI including rotator cuff arthropathy, proximal migra-tion of humerus, rotator cuff tear size, fatty infiltration,and muscle atrophy were gathered preoperatively. Allparameters were collected twice by each of the two inde-pendent observers, in order to compare intra- andinter-observer reliability. However, we used the datafrom experienced orthopaedist only to analyze for theprognostic factors.The arthroscopic surgery was performed by three experi-

enced surgeons after the patient was sedated with generalanesthesia and set in the beach chair position. In cases ofadhesive capsulitis, the capsular release and manipulationwas done in the same setting. Acromioplasty was per-formed if an acromial spur was defined. After the tear sizewas measured with a probe to confirm the size measuredfrom the MRI, the surgeon released any adhesion sur-rounding the tendon. The tendon then was mobilized inorder to cover its footprint as much as possible. If neces-sary, the interval sliding technique was also performed.The transosseous-equivalent double-row technique was

chosen to be the first line of treatment. If the double-rowtechnique were not feasible, a single-row technique or par-tial functional repair technique was chosen instead.Complete rotator cuff repair was defined as the tendon

covering up at least 50% of the anatomical footprint, whilepartial repair was defined as the tendon covering up lessthan 50% of the original footprint (Fig. 1) [12, 15]. All pa-tients were underwent passive motion exercises on thefirst post-operative days. An arm sling was applied for 6weeks. A progressive active-assisted passive motion exer-cises were commenced at the third to fourth week formuscle strengthening. This research was approved by Re-search and Ethic Committee from Queen Savang Vadhanamemorial hospital.

Radiographic parametersPre-operative plain radiograph focused on three param-eter measurements to evaluate rotator cuff arthropathywhich include osteoarthritic change, acromiohumeraldistance, and inferior glenohumeral distance. Furtherpre-operative MRI evaluation focused on mediolateral(ML) tear size, anteroposterior (AP) tear size, tendon re-traction, fatty infiltration and muscle atrophy. Radio-graphic parameters were evaluated by one experiencedorthopedist and one senior orthopedic resident. A 1.5-TMRI (Siemens Healthcare, Germany) was used in thisstudy. All measurements were made on a PACS worksta-tion using Agfa IMPAX 6 (Waterloo, Canada) softwaretechnology—this software uses data within the DICOM(NEMA, VA, USA) header on all MRI (whether fromour institution or from outside institutions) to allow ref-erenced measurements to be made.

Rotator cuff ArthropathyThe rotator cuff arthropathy was graded by Hamada’sclassification [5], which showed good inter- andintra-observer reliability from a previous study [16]. Thestudied population were classified into 5 grades: grade 1,the AHI is ≥6 mm; grade 2, the AHI is ≤5 mm; grade 3,AHI is ≤5 mm with acetabulization of the acromion;grade 4, glenohumeral narrowing; and grade 5, humeralhead collapse [7].Superior Migration of the Humeral Head; Acromion-

Humeral Interval (AHI) & Inferior Glenohumeral Dis-tance (IGHD).Acromion-humeral interval (AHI) was one of the par-

ameter to evaluate superior migration of the humeral head(SMHH). It was measured on an anteroposterior view ofthe plain radiograph, as the distance between inferiorborder of acromion and superior aspect of the humeralhead. Inferior glenohumeral distance (IGHD) was anotherparameter to evaluate SMHH measuring the distance be-tween inferior glenoid margin and the inferior humeral

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 2 of 9

head margin at just the intersection of the humeral headand the humeral neck [12] (Fig. 2).

Tear sizeTear size of supraspinatus tendon was collected fromT2-weighted MRI in both coronal oblique and sagittal

oblique views which gave mediolateral (ML) tear size andanteroposterior (AP) tear size respectively. ML tear sizewas a straight distance from the tendon edge to the lateralcortex of the greater tuberosity in the coronal oblique cut,which the tendon most retracted medially. AP tear sizewas a longest straight distance from the anterior tendonedge to the posterior tendon edge in the sagittal obliquecut [17]. The frayed tissues at the tip of tendon edge werenot included in any measurement (Fig. 3).

Tendon retractionFrom the coronal oblique view of T2-weight MRI, weuse the slice which showed the most medial tendon re-traction. The tendon retraction was classified using theclassification described by Patte [18]. In stage 1, the ten-don stump was close to the bony insertion. In stage 2,the tendon stump was retracted and lied at the level ofhumeral head between the footprint and glenoid. Instage 3, the tendon stump was seen at the level of theglenoid or beyond.

Fatty infiltrationUsing the sagittal oblique view of T1-weight MRI, fattyinfiltration in all rotator cuff muscles were classified bythe modified Goutallier staging system described byFuch et al. [19]. Based on the lateral image on which thescapular spine was in contact with scapular body(Y-shaped view), the patients were classified into 5grades: grade 0, normal muscle without fatty streak;grade 1, some fatty streaks in the muscle; grade 2, fattyinfiltration is present, but more muscle than fat; grade 3,equal amount of fat and muscle; and grade 4, more fatthan muscle.

Muscle atrophyWe used the sagittal oblique view of T1-weighted MRIto evaluate the rotator cuff muscle atrophy by selecting

Fig. 1 a After arthroscopic rotator cuff repair, the tendon covered more than 50% of the anatomical footprint in completely repaired group. b Inan partially repaired group, the tendon covered less than 50% of the anatomical footprint. (GT = greater tuberosity)

Fig. 2 The superior migration of humeral head was defined withtwo methods of measurement. Acromioclavicular interval (arrow)was measured between the inferior border of acromion and thesuperior aspect of humeral head. Inferior glenohumeal distance(arrow head) was measured between the inferior glenoid marginand the inferior humeral head margin at just the intersection ofhumeral head and the humeral neck

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 3 of 9

the most lateral image on which the scapular spine wasin contact with scapular body (Y-shaped view). First, theZanetti’s tangent line extended from the superior aspectof the coracoid to the superior border of the scapularspine and the other two lines were extended from eachscapular process to the tip of scapular body (Fig. 4).These lines divided muscle atrophy into 4 groups: noneor no muscle atrophy, the muscle crossing over the tan-gent line; mild, border of the muscle touching the tan-gent line; moderate, a concave curve of the musclebeside the tangent line; and severe, the muscle bundleatrophy nearly touching the scapular spine and scapularbody [20].

Statistical analysisAs the data were not normally distributed, we selected theChi-square test to compare proportions of categorical vari-ables between complete and partial repair group. These var-iables included gender, traumatic event and tendoninvolvement. A Mann-Whitney U test was used to comparebetween continuous data variables in the complete and par-tial repair groups. These include age, duration of symptom,AHI and IGHD, ML and AP tear size, grading of fatty infil-tration and muscle atrophy. Pearson-correlation analysisand Spearman rank-order correlation took a role in deter-mining association between categorical and continuous var-iables, respectively. These variables, which determined theindependent factors affecting reparability, were analyzed ina stepwise multivariate logistic regression analysis. Wecreated the receiver operating characteristic (ROC)curve of significant predicting factors which correlated

Fig. 3 a From magnetic resonance imaging, mediolateral tear size was measured between the tendon edge and the lateral cortex of greatertuberosity in the coronal oblique cut, which the tendon most retracted medially. b Anteroposterior tear size was measured between the anteriortendon edge to the posterior tendon edge in the sagittal oblique cut, which showed the longest distance

Fig. 4 From the sagittal oblique view of T1-weighted magneticresonance imaging, the muscle atrophy were evaluated in the mostlateral image on which the scapular spine was in contact withscapular body (Y - shaped view). The Zanetti’s tangent line extendedfrom superior aspect of the coracoid to the superior border of thescapular spine, then the other two lines were extended from eachscapular process to tip of scapular body

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 4 of 9

with reparability of rotator cuff tendons. The cut-offlevels of each factor were defined from the ROC curve.The likelihood ratios of each factor were then calcu-lated to weight the importance to rotator cuff tendonreparability.Intra-observer and inter-observer reliabilities were

evaluated with the Kappa analysis and intraclass correl-ation coefficient for categorical and continuous datarespectively.Statistical significant was set at p-value < 0.05. We use

the IBM SPSS statistics software version 20 for all statis-tical analysis.

ResultsSixty-six patients (31 women and 35 men, mean age 60years, range 25–84 years) were included in the analysis.Eleven large-sized rotator cuff tears (16.67%) andfifty-five massive rotator cuff tears (83.33%) were definedfrom MRI and all confirmed under arthroscopic examin-ation. The mean duration between MRI and surgery was5.5 weeks. Overall, 54 and 12 patients underwentcomplete and partial repair, respectively. Demographicdata comparing between each group with respect to age,gender, traumatic event, duration of symptom, durationbetween MRI to surgery, rotator cuff arthropathy, AHIand IGHD, ML and AP tear size, grading of fatty infiltra-tion, and muscle atrophy as showed in Table 1. Thereare nine parameters; age, ML tear size, AP tear size, ro-tator cuff arthropathy, AHI, IGHD, fatty infiltration andmuscle atrophy of the supraspinatus muscle, and onlyfatty infiltration of the infraspinatus muscle which showa significant correlation with partially repaired cases.The mean age was 59.3 years in the complete repair

group and 67.7 years in the partial repair group. This re-vealed a significant age difference in the partial repairgroup(p = .015). Both ML and AP tear size also showed acorrelation with partial repair with p-value of .001 and.002 respectively. Rotator cuff arthropathy, which wasgraded by the Hamada classification, AHI and IGHD wereall correspondingly related to partial repair (p = .004,p = .003, p = .036) Fatty infiltration and muscle atro-phy of supraspinatus muscle was significantly associ-ated with partial repair(p < .001, p = .001). Fattyinfiltration of infraspinatus was correlated with par-tial repair(p = .031).The ROC curves were further analyzed on these nine

significant predicting factors, age, ML tear size, AP tearsize, rotator cuff arthropathy, AHI, IGHD, fatty infiltra-tion, and muscle atrophy of the supraspinatus muscle,and fatty infiltration of the infraspinatus muscle tosearch for each proper cut-off point that gave optimalsensitivity, specificity and likelihood ratio. The resultswere as follows; age of 65 years old, 22 mm of AP tearsize, 36 mm of ML tear size, class 2 of Hamada’s rotator

cuff arthropathy, AHI of 6 mm, IGHD of 5mm, stage 3of fatty infiltration of supraspinatus muscle and stage 1of fatty infiltration of infraspinatus muscle, and mild de-gree of supraspinatus muscle atrophy gave the optimalsensitivity and specificity (Table 2). The post-hoc ana-lysis was performed to define the power of each signifi-cant predicting factor. The power of each factor rangesfrom 70.2 to 98.2%.We have found there were only three independent fac-

tors that significantly affected the reparability after mul-tivariated logistic regression analysis. These were age of65 years or older (p = .004), acromiohumeral index ofless than 6 mm. (p = .007), and anteromedial tear size ofmore than 22mm. (p = .026) (Table 3).The intra-observer reliability ranged from 60 to 100%

with average 85.69%. The inter-observer reliabilityranged from 43.33 to 100% with an average of 74.5%.Both results were considered as good to excellent.

DiscussionArthroscopic surgery is now becoming a more for popu-lar option for rotator cuff repair as it shows good to ex-cellent results. However, large-sized and massive rotatorcuff tears remain a challenge situation as there is a highre-rupture rate [11, 21–23]. Previously, Burkhart has de-scribed a concept of functional partial repair which re-duces pain and improves shoulder function [24–27].Furthermore, recent studies showed that partial repairprovide inferior functional outcomes, shoulder mobilityand strength compared to the anatomical repair [8].Tendon mobility to cover the footprint could only bedemonstrated under the operation. Nonetheless, variousoptions of arthroscopic and open surgery were per-formed for the partially repaired rotator cuff tears ran-ging from arthroscopic debridement to reversed totalshoulder replacement as described in the introduction.The transosseous-equivalent double-row technique is

widely used in arthroscopic rotator cuff repairs as it hasbeen shown in several biomechanics studies to have ahigher tensile strength than a conventional single-rowtechnique [28, 29]. The previous meta-analysis studiesshowed that double-row repair had a more completehealing rate, less re-rupture rate, better range of motion,and improved clinical outcomes, especially in large-sizedand massive rotator cuff tears [30–32]. Hence, in thisstudy, we intended to arthroscopically repair rotator cufftendons with transosseous-equivalent double-row tech-nique to completely cover the anatomical footprint asmuch as possible. In case the tendon is unable to be fullymobilized, single-row repair or functional partial repairwas chosen instead.Other recent studies have explored the clinical and sur-

gical factors in predicting reparability of large and massiverotator cuff tears. A study by Holtby and Razmjou showed

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 5 of 9

Table 1 Descriptive data of 66 patients (Complete Repair vs Partial Repair group)

Predictive factors Complete Repair (n = 54) Partial Repair (n = 12) p-value

Age 59.31(range 25–84) 67.75(range 47–81) .015

Male 27 (50%) 8 (66.67%) 0.295

Duration of onset (weeks) 22.24(range 3–98) 27.54(range 5–104) 0.634

Traumatic event 31 (57.41%) 5 (41.67%) 0.322

Rotator cuff arthropathy

• Grade 0 15 (27.78%) 0 (0%) 0.004

• Grade 1 33 (61.11%) 5 (41.67%)

• Grade 2 3 (5.56%) 5 (41.67%)

• Grade 3 2 (3.70%) 1 (8.33%)

• Grade 4 1 (1.85%) 1 (8.33%)

• Grade 5 0 (0%) 0 (0%)

AHI 7.92 (range 0–15.36) 5.14 (range 1.30–10.21) 0.003

IGHD 3.85 (range (−4.3)-14.36) 6.62 (range (−3.35)-10.85) 0.036

Mediolateral tear size 28.48 (range11.74–48.3) 39.75 (range28–48.27) .001

Anteroposterior tear size 19.94 (range0–44.56) 30.76 (range10–47.3) .002

Fatty infiltration (Goutallier staging)

• Supraspinatus muscle

ο Stage 0 33 (61.11%) 1 (8.33%) <.001

ο Stage 1 10 (18.52%) 3 (25%)

ο Stage 2 5 (9.26%) 2 (16.67%)

ο Stage 3 3 (5.56%) 5 (41.67%)

ο Stage 4 3 (5.56%) 1 (8.33%)

• Infraspinatus muscle

ο Stage 0 35 (64.81%) 3 (25%) .031

ο Stage 1 5 (9.26%) 3 (25%)

ο Stage 2 5 (9.26%) 2 (16.67%)

ο Stage 3 4 (7.41%) 3 (25%)

ο Stage 4 5 (9.26%) 1 (8.33%)

• Subscapularis muscle

ο Stage 0 36 (66.67%) 5 (41.67%) .199

ο Stage 1 9 (16.67%) 4 (33.33%)

ο Stage 2 3 (5.56%) 2 (16.67%)

ο Stage 3 2 (3.70%) 1 (8.33%)

ο Stage 4 4 (7.41%) 0 (0%)

• Teres minor muscle

ο Stage 0 40 (74.07%) 8 (66.67%) .701

ο Stage 1 9 (16.67%) 3 (25%)

ο Stage 2 3 (5.56%) 1 (8.33%)

ο Stage 3 0 (0%) 0 (0%)

ο Stage 4 2 (3.70%) 0 (0%)

Muscle atrophy

• Supraspinatus muscle

ο None 33 (61.11%) 0 (0%) .001

ο Mild 11 (20.37%) 7 (58.33%)

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 6 of 9

that U-shaped tears, tear size, and tendon quality definedintraoperatively correlated with reparability. On a con-trary, there is no pre-operative factors that can predict thereparability of large and massive rotator cuff tears [8].Dwyer et al. studied the association between pre-operativeMRI and reparability of large and massive rotator cufftears. The results were quite similar with previous studies,demonstrating mediolateral tear size, tendon retraction tothe glenoid level, muscle atrophy, advanced fatty infiltra-tion and superior migration of humeral head were associ-ated with partial repairs [15].Our study included clinical, radiographic and MRI pa-

rameters to analyze the correlation with reparability oflarge-sized and massive rotator cuff tears. The rotatorcuff tear is a degenerative disease as Pecora et al. dem-onstrated that age is associated with clinical outcomes

after rotator cuff repair [33]. From our study, age is oneof the factors that correlates with rotator cuff reparabil-ity. The tendon tissue quality, which differs in variousage groups, is also likely an important factor. The mech-anism of rotator cuff tear differs among different agegroups. Younger patients often presented with acutetraumatic tears, while older patients often presentedwith chronic degenerative tears. It is found that the like-lihood ratio of partial repair is 2.81 in patients 65 yearsor older. In traumatic rotator cuff tears, there is moreadhesion and scar in a subacromial space. Although theadhesion limits the tendon mobility and makes a repairmore difficult, a traumatic event does not correlate withreparability in this study.Our study showed that the onset of the symptom is not

correlated with AHI, fatty infiltration, muscle atrophy, and

Table 1 Descriptive data of 66 patients (Complete Repair vs Partial Repair group) (Continued)

Predictive factors Complete Repair (n = 54) Partial Repair (n = 12) p-value

ο Moderate 7 (12.96%) 5 (41.67%)

ο Severe 3 (5.56%) 0 (0%)

• Infraspinatus muscle

ο None 44 (81.48%) 7 (58.33%) .163

ο Mild 3 (5.56%) 4 (33.33%)

ο Moderate 2 (3.73%) 0 (0%)

ο Severe 5 (9.26%) 1 (8.33%)

• Subscapularis muscle

ο None 36 (87.72%) 11 (91.67%) .639

ο Mild 3 (6.98%) 0 (0%)

ο Moderate 0 (0%) 1 (8.33%)

ο Severe 4 (9.30%) 0 (0%)

• Teres minor muscle

ο None 47 (87.04%) 12 (100%) .502

ο Mild 3 (5.56%) 0 (0%)

ο Moderate 0 (0%) 0 (0%)

ο Severe 4 (7.41%) 0 (0%)

Table 2 Results from Receiver Operating Characteristic curve analysis of predicting factors of arthroscopic reparability for large-sizedand massive rotator cuff tears

Factor Cut point Sensitivity (SD) Specificity (SD) Likelihood ratio P-value

Age ≥ 65 years old 0.83 (0.04) 0.70 (0.03) 2.81 0.015

ML tear size ≥ 36mm 0.91 (0.02) 0.74 (0.03) 3.51 0.001

AP tear size ≥ 22mm 0.75 (0.05) 0.70 (0.03) 2.65 0.002

Rotator cuff arthropathy ≥ stage 2 0.58 (0.07) 0.89 (0.01) 5.25 0.004

AHI ≤ 6 mm 0.64 (0.07) 0.78 (0.02) 2.86 0.003

IGHD ≥ 5 mm 0.73 (0.06) 0.60 (0.03) 1.79 0.036

Supraspinatus fatty infiltration ≥ stage 3 0.50 (0.07) 0.89 (0.01) 4.50 < 0.001

Infraspinatus fatty infiltration ≥ stage 1 0.75 (0.05) 0.64 (0.03) 2.79 0.031

Supraspinatus muscle atrophy mild to severe 1.00 (N.A.) 0.60 (0.02) 2.52 0.001

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 7 of 9

rotator cuff arthropathy, all of which are the parameters ofchronicity. One explanation might be the recall bias fromthe patients. Nevertheless, this aspect supports that rota-tor cuff tears are asymptomatic may be silent for certainperiod of time before they were clinically detected.This analysis showed stage 2 or more of Hamada’s

classification is related with high risk of partial repair.Our study also found that in stage 2 of Hamada’s classifi-cation with less than 5 mm of AHI, the superior migra-tion of humeral head significantly correlated with partialrepair, similar to prior studies as well as IGHD that wasassociated with the reparability [15]. Regardless, ourstudy shows AHI 6mm or more and IGHD 5mm ormore correlate with partial repair of rotator cuff tearswith the likelihood ratio of partial repair of 2.86 and1.79, respectively.For MRI parameters, there were previous studies which

demonstrated the association between tear sizes and par-tial repair [34]. Yoo et al. reported the sagittal tear size ofequal to or greater than 32mm and the coronal tear sizeof equal to or longer than 31mm correlated with an in-ability to achieve satisfactory anatomical rotator cuff re-pair [21]. Sugihara et al. showed that primary rotator cuffrepair was usually not feasible when both the length andwidth of tear are equal to or longer than 40mm [34]. Ourstudy demonstrated a mediolateral tear size of 36mm ormore and anteroposterior tear size 22mm or more corre-lated with partial repair of rotator cuff tears with likeli-hood ratio of partial repair of 3.51 and 2.65, respectively.Tendon retraction, a parameter that can resemble the tearsize, but more simple to use and more reliable. On a con-trary, it is not significantly correlated with reparability.Advanced fatty infiltration, which was described by Gou-

tallier et al., correlates with an inability to achieve completerotator cuff repair [15, 21]. Our study demonstrates ad-vanced fatty infiltration of supraspinatus muscle (stage 3 ormore) and mild degree of infraspinatus muscle (stage 1 ormore) correlated with an unsuccessful rotator cuff repairwith the the likelihood ratio of partial repair of 4.50 and2.79, respectively. The classification of muscle atrophy,which Thomazeau et al. described, is widely used [35].From our study, the muscle atrophy of stage I or greater,where the muscle mass is just at the same level of a tangentline or below, correlated with partial repair with the likeli-hood ratio of 2.52.

The limitation of this study is a low number in the par-tial repair group. Comparing to prior studies [8, 15, 34],the rate of partially repaired rotator cuff tear in our studyis quite different, despite the fact that we use the same cri-teria of reparability. There was a broad range of age group,which may impact the tissue quality and probability oftraumatic rotator cuff tear. However, the post-hoc analysisshowed the powers of each significant predicting factorrange from 70.2 to 98.2% despite only 12 patients in par-tial repair group.

ConclusionsReparability of large-to-massive rotator cuff tears correl-ate with the following factors; age, AP tear size, ML tearsize, rotator cuff arthropathy, AHI, IGHD, fatty infiltra-tion of supraspinatus and infraspinatus muscles, and at-rophy of supraspinatus muscle.

AbbreviationsAHI: Acromiohumeral interval; AP: Anteroposterior; IGHD: Inferiorglenohumeral distance; LR: Likelihood ratio; ML: Mediolateral; MRI: Magneticresonance imaging; ROC: receiver operating characteristic; SMHH: Superiormigration of humeral head

AcknowledgementsNot applicable

FundingThere was no external funding or grant for this study.

Availability of data and materialsThe datasets used and analyzed during the current study are available fromthe corresponding author on reasonable request.

Authors’ contributionsAll authors were involved in the design of the study and in the analysis anddiscussion of the results. VK performed most of the data analysis, and wrotethe first draft of the manuscript. NP collected the data, supervised the entirestudy process and helped to write and review the manuscript. TL and VUparticipated in the data collection. CY participated in the study design andhelped to analyse the data. All authors read and approved the final manuscript.

Ethics approval and consent to participateThis study was approved by research and ethic committee from Queensavang vadhana memorial hospital, and written informed consent wasobtained from all the participants before the study was conducted.

Consent for publicationNot applicable

Competing interestsThe authors declare that they have no competing interests.

Table 3 Significant factors, in multiple logistic regression models, associated with arthroscopic reparability of large and massiverotator cuff tear

Factor Coeff (Std. err) OR 95% CI P-value

Lower Bound Upper Bound

Age≥ 65 years old 0.333 1.395 0.085 0.429 0.004

AHI≤ 6 mm 0.297 1.346 0.067 0.417 0.007

AP tear size ≥22 mm 0.243 1.275 0.023 0.359 0.026

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 8 of 9

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Received: 16 May 2018 Accepted: 3 March 2019

References1. Reilly P, Macleod I, Macfarlane R, Windley J, Emery RJ. Dead men and

radiologists don't lie: a review of cadaveric and radiological studies ofrotator cuff tear prevalence. Ann R Coll Surg Engl. 2006;88(2):116–21.

2. Yamamoto A, Takagishi K, Osawa T, Yanagawa T, Nakajima D, Shitara H, etal. Prevalence and risk factors of a rotator cuff tear in the generalpopulation. J Shoulder Elb Surg. 2010;19(1):116–20.

3. Warner JJ. Management of massive irreparable rotator cuff tears: the role oftendon transfer. Instr Course Lect. 2001;50:63–71.

4. Sonnabend DH, Watson EM. Structural factors affecting the outcome ofrotator cuff repair. J Shoulder Elb Surg. 2002;11(3):212–8.

5. Hamada K, Yamanaka K, Uchiyama Y, Mikasa T, Mikasa M. A radiographicclassification of massive rotator cuff tear arthritis. Clin Orthop Relat Res.2011;469(9):2452–60.

6. Kuzel BR, Grindel S, Papandrea R, Ziegler D. Fatty infiltration and rotator cuffatrophy. J Am Acad Orthop Surg. 2013;21(10):613–23.

7. Le BT, Wu XL, Lam PH, Murrell GA. Factors predicting rotator cuff retears: ananalysis of 1000 consecutive rotator cuff repairs. Am J Sports Med. 2014;42(5):1134–42.

8. Holtby R, Razmjou H. Relationship between clinical and surgical findingsand reparability of large and massive rotator cuff tears: a longitudinal study.BMC Musculoskelet Disord. 2014;15:180.

9. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC. Fatty muscledegeneration in cuff ruptures. Pre- and postoperative evaluation by CT scan.Clin Orthop Relat Res. 1994;304:78–83.

10. Goutallier D, Postel JM, Gleyze P, Leguilloux P, Van Driessche S. Influence ofcuff muscle fatty degeneration on anatomic and functional outcomes aftersimple suture of full-thickness tears. J Shoulder Elb Surg. 2003;12(6):550–4.

11. Mellado JM, Calmet J, Olona M, Esteve C, Camins A, Perez Del Palomar L, etal. Surgically repaired massive rotator cuff tears: MRI of tendon integrity,muscle fatty degeneration, and muscle atrophy correlated with intraoperativeand clinical findings. AJR Am J Roentgenol. 2005;184(5):1456–63.

12. Koh KH, Lim TK, Park YE, Lee SW, Park WH, Yoo JC. Preoperative factorsaffecting footprint coverage in rotator cuff repair. Am J Sports Med. 2014;42(4):869–76.

13. Millstein ES, Snyder SJ. Arthroscopic evaluation and management of rotatorcuff tears. Orthop Clin North Am. 2003;34(4):507–20.

14. Aliprandi A, Messina C, Arrigoni P, Bandirali M, Di Leo G, Longo S, et al.Reporting rotator cuff tears on magnetic resonance arthrography using theSnyder's arthroscopic classification. World J Radiol. 2017;9(3):126–33.

15. Dwyer T, Razmjou H, Henry P, Gosselin-Fournier S, Holtby R. Associationbetween pre-operative magnetic resonance imaging and reparability oflarge and massive rotator cuff tears. Knee Surg Sports Traumatol Arthrosc.2015;23(2):415–22.

16. Kappe T, Cakir B, Reichel H, Elsharkawi M. Reliability of radiologicclassification for cuff tear arthropathy. J Shoulder Elb Surg. 2011;20(4):543–7.

17. Davidson JF, Burkhart SS, Richards DP, Campbell SE. Use of preoperativemagnetic resonance imaging to predict rotator cuff tear pattern andmethod of repair. Arthroscopy. 2005;21(12):1428.

18. Patte D. Classification of rotator cuff lesions. Clin Orthop Relat Res. 1990;254:81–6.

19. Fuchs B, Weishaupt D, Zanetti M, Hodler J, Gerber C. Fatty degeneration ofthe muscles of the rotator cuff: assessment by computed tomographyversus magnetic resonance imaging. J Shoulder Elb Surg. 1999;8(6):599–605.

20. Warner JJ, Higgins L, IMt P, Dowdy P. Diagnosis and treatment ofanterosuperior rotator cuff tears. J Shoulder Elb Surg. 2001;10(1):37–46.

21. Yoo JC, Ahn JH, Koh KH, Lim KS. Rotator cuff integrity after arthroscopicrepair for large tears with less-than-optimal footprint coverage. Arthroscopy.2009;25(10):1093–100.

22. Porcellini G, Castagna A, Cesari E, Merolla G, Pellegrini A, Paladini P. Partialrepair of irreparable supraspinatus tendon tears: clinical and radiographicevaluations at long-term follow-up. J Shoulder Elb Surg. 2011;20(7):1170–7.

23. Chung SW, Kim JY, Yoon JP, Lyu SH, Rhee SM, Oh SB. Arthroscopic repair ofpartial-thickness and small full-thickness rotator cuff tears: tendon quality asa prognostic factor for repair integrity. Am J Sports Med. 2015;43(3):588–96.

24. Burkhart SS. Fluoroscopic comparison of kinematic patterns in massiverotator cuff tears. A suspension bridge model. Clin Orthop Relat Res. 1992;284:144–52.

25. Burkhart SS, Esch JC, Jolson RS. The rotator crescent and rotator cable: ananatomic description of the shoulder's "suspension bridge". Arthroscopy.1993;9(6):611–6.

26. Burkhart SS. Reconciling the paradox of rotator cuff repair versusdebridement: a unified biomechanical rationale for the treatment of rotatorcuff tears. Arthroscopy. 1994;10(1):4–19.

27. Burkhart SS, Lo IK. Arthroscopic rotator cuff repair. J Am Acad Orthop Surg.2006;14(6):333–46.

28. Snyder SJ. Technique of arthroscopic rotator cuff repair using implantable 4-mm Revo suture anchors, suture shuttle relays, and no. 2 nonabsorbablemattress sutures. Orthop Clin North Am. 1997;28(2):267–75.

29. Smith CD, Alexander S, Hill AM, Huijsmans PE, Bull AM, Amis AA, et al. Abiomechanical comparison of single and double-row fixation inarthroscopic rotator cuff repair. J Bone Joint Surg Am. 2006;88(11):2425–31.

30. Prasathaporn N, Kuptniratsaikul S, Kongrukgreatiyos K. Single-row repairversus double-row repair of full-thickness rotator cuff tears. Arthroscopy.2011;27(7):978–85.

31. Zhang Q, Ge H, Zhou J, Yuan C, Chen K, Cheng B. Single-row or double-rowfixation technique for full-thickness rotator cuff tears: a meta-analysis. PLoSOne. 2013;8(7):e68515.

32. Xu C, Zhao J, Li D. Meta-analysis comparing single-row and double-rowrepair techniques in the arthroscopic treatment of rotator cuff tears. JShoulder Elb Surg. 2014;23(2):182–8.

33. Pecora JO, Malavolta EA, Assuncao JH, Gracitelli ME, Martins JP, Ferreira AAJr. Prognostic factors for clinical outcomes after rotator cuff repair. ActaOrtop Bras. 2015;23(3):146–9.

34. Sugihara T, Nakagawa T, Tsuchiya M, Ishizuki M. Prediction of primaryreparability of massive tears of the rotator cuff on preoperative magneticresonance imaging. J Shoulder Elb Surg. 2003;12(3):222–5.

35. Thomazeau H, Boukobza E, Morcet N, Chaperon J, Langlais F. Prediction ofrotator cuff repair results by magnetic resonance imaging. Clin Orthop RelatRes. 1997;344:275–83.

Kuptniratsaikul et al. BMC Musculoskeletal Disorders (2019) 20:111 Page 9 of 9